algorithmic modeling for Rhino
Grasshopper (and indeed Rhino) can be performance critical applications since both potentially deal with large amounts of data and computations. Although we aim to make our software runnable on low-end, over-the-counter computers, you may still run into serious performance issues. We have no strict recommendation or requirements, but here are the basic rules when it comes to picking hardware for Rhino and Grasshopper:
Some further points to take into account:
* This may change in the future, but not the foreseeable future.
Regarding Memory. Does RAM frequency have much of an impact on the performance of Grasshopper or is it pretty much down to the GB? From reading performance tests for games it seems the difference provided by higher frequencies is negligible for that usage, equating to a percent or two. How about Grasshopper?
As an example, would one notice much of a difference in performance between an 8GB stick of 2133MHz DDR4 RAM and a 2400MHz stick? How about a 3200MHz stick? Assume Grasshopper is running on a 64bit system.
A follow-up question: Faster RAM frequencies are more expensive than slower. So, constrained by a limited budget, would it be better to upgrade to 8GB of 3200MHz DDR RAM (replacing an existing 8GB 2133MHz stick), or 16GB of 2133MHz DDR RAM (Adding an additional 8GB 2133MHz stick to the existing one)? From what I understand adding faster sticks to slow ones is kind of pointless as they are limited to the speed of the slowest stick, so I have not included that as an option.
I do not know, but I doubt the difference is large. If you're investing in memory you're better off getting larger L1/L2/L3 caches.
Right. I assumed so. I chose to upgrade to 16GB of 2133MHz DDR4. It certainly was a noticeable improvement, but I am already maxing out the RAM again with certain operations, particularly complex meshes. Looking into renting processing power to get over the computational hump for complex operations.